Suicide is an urgent and growing public health crisis. It was the tenth leading cause of death in the United States in 2015, with over 44,000 deaths, according to the CDC. Studying suicide, and including suicidal or potentially suicidal individuals in clinical research, is an important way to gain valuable data that can advance prevention efforts. Researchers and IRBs may tend to exclude suicidal individuals from research in order to avoid potential risks; however, in order for the research to be scientifically and clinically valuable, it is important that suicidal individuals are not excluded from research unnecessarily.
In October 2017, PRIM&R hosted a webinar, Suicidal Individuals in Research: Ethical and Safety Considerations, to offer considerations and strategies for IRBs involved in reviewing research that includes individuals that may be suicidal. Speaker Galia Siegel, PhD, clinical trials program coordinator in the Office of Clinical Research at the National Institutes of Mental Health (NIMH), addressed recently released guidance* on this topic, giving attention to areas most relevant to IRBs, including informed consent, study design, and safety monitoring and reporting. Barbara Stanley, PhD, director of the Center on Intervention for Prevention of Suicide and research scientist in the Department of Neuroscience at New York State Psychiatric Institute, emphasized the importance of ensuring suicidal individuals are not unnecessarily excluded from research, and offered strategies for mitigating risk through proper implementation of safety measures.
After the webinar, Dr. Siegel and Dr. Stanley responded to some of the attendee questions time didn’t permit us to address live. We’re pleased to share those responses with the readers of Ampersand.
If you include suicidal subjects, often they will need to be analyzed as a separate stratum, which will increase N to achieve statistical significance for both suicidal and non-suicidal subjects, unless the numbers of subjects are increased considerably. How do you deal with this?
Barbara Stanley (BS): Suicidal participants do not necessarily have to be analyzed as a separate group, just as you do not need to analyze participants endorsing separate symptoms of a diagnosis separately. So, for example, if you are conducting a study of major depressive disorder, you do not need to analyze all patients with insomnia separately from those without insomnia. You would only do this analysis if you are interested in the differential effects of the treatment on those with and without insomnia. However, if you are doing a study of major depression and you systematically exclude suicidal individuals, generalizability is compromised.
What is the proper way to conduct clinical management of potentially suicidal individuals in research that is done remotely? How do you make it work, for example, with subjects in rural communities who may be remote from the study site?
Galia Siegel (GS): There are many potential options to consider depending on your study design and methods, as well as the risks presented by your study participants. Risk assessments are often conducted remotely, via telephone or other forms of electronic communication. The frequency and method of those assessments should be laid out in your protocol, as well as the level of training of the study staff conducting follow up assessments. Your protocol should also discuss what will trigger the need for clinical follow up, and what the follow up will be for specified levels of risk. Some clinical management options include: 1) as part of study participation, develop a safety plan with participants based on the resources that are available to the participants locally and; 2) establish a relationship with a crisis line and set up a protocol for a warm transfer for at-risk participants; 3) obtain permission to inform a person’s local care provider of their level of risk; 4) have a safety officer as part of the study staff available to conduct a thorough risk assessment with the person on the phone once a heightened level of risk has been identified; 5) inform participants clearly at the outset of the study (and as part the informed consent process) what the risk assessment and clinical management plans will be for study participants, and any limitations present for remote participants. The most appropriate way to provide remote clinical management for a specific study with potentially suicidal participants will depend on the specifics of the study. Consulting with investigators with experience doing this kind of research is an effective way to come up with a set of potential options for the study team to consider.
What do you recommend as a best practice for suicide risk management in online research with individuals at risk for suicide (when the research is not an intervention study – e.g., to gather data on psychometrics of a suicide risk assessment tool)? Is provision of national referrals based on an algorithm appropriate/sufficient? Does the study team need to be available 24/7? Often the research in this area that I review as an IRB member is completed without asking for identifiable information in order to protect subjects’ privacy and to reduce the risk of negative consequences due to an unintended data breach.
BS: Study team availability 24/7 is not necessary. It is a good idea to post hotline and referral information for everyone. If a risk assessment tool is being validated, it is hard to know how it could be used for an algorithm. However, if a person endorses items that we know to be indicators of high risk (e.g. suicidal ideation with intent and plan), one could consider having a flag that says something like, “You seem to be struggling…” and then make a suggestion about how they can reach out.
Do you have any recommendations about professional qualifications for those who would be making suicidal ideation and behavior (SIB) assessments?
GS: To address this question, it’s necessary to think about multiple aspects of SIB assessments. One is what training and qualifications do study staff need to ask the questions involved in an SIB assessment and do an initial triage if someone is identified to be at risk. Another aspect is what training and qualifications study staff need to provide appropriate clinical management for someone with identified risk. Some study teams have staff with clinical training (masters level clinician, or a PhD or MD) conducting SIB assessment so that they can both ask the questions, and they also have the clinical training and experience to manage the clinical follow up in cases of identified risk. Other studies have individuals with a bachelor’s degree or individuals completing a bachelor’s or master’s programs, working as research assistants. In these cases, the research assistants are provided with training to be able to conduct the SIB assessments and identify individuals who need to be triaged to a study staff person or designated non-study clinical staff who have the qualifications to respond to someone who has endorsed suicide risk.
Are you in favor of prospective monitoring of suicidal thoughts and behaviors in essentially all clinical trials as part of broad suicide prevention initiatives?
GS: The NIMH has identified suicidal ideation as a research priority and is very supportive of clinical trials that are not suicide-specific, including assessments of suicidal ideation and behavior. Nevertheless, the decision of whether to include prospective monitoring of suicide-related events in a specific study will depend on the study team. Research teams would need the resources and capacity to do work as part of protocol development and design to determine what monitoring questions would be included, as well as to address staffing and clinical management questions. For example, will staff administer questionnaires? If so, what training do they need, and what clinical back up is needed? If assessments will be conducted electronically, what procedures need to be in place to monitor and respond to the risk data?
Do you obtain a certificate of confidentiality for all of your studies?
GS: The NIH has recently implemented a policy of automatically providing certificates of confidentiality for all NIH-funded research that collects or uses identifiable, sensitive information. As a result, all NIMH funded studies now have certificates of confidentiality.
What protection procedures might be used by researchers who only interact with subjects once or in a limited timeframe, which would make removal from the study and monitoring subjects over time difficult?
GS: A key issue in developing appropriate monitoring and safety practices is identified in this question, specifically that the appropriate safety procedures for a given study will depend on the specifics of the participants’ involvement. If a participant’s study involvement includes one visit, then any risk assessment and clinical management might take place at that visit without any follow up by the study team. For participants who are suicidal or at heightened risk for suicide, providing crisis line or referral information, and/or developing a safety plan could be considered as potential activities for that visit. Similarly, for studies completed in a limited time frame (i.e., a week or a month), the protections might include the same ones as those mentioned above, with risk assessments being conducted at a frequency appropriate both to the risk level of the participants and to the design of the study. Consulting with researchers who have experience conducting brief studies with at-risk populations might generate some useful ideas.
Since completed suicides may be of low incidence, can you comment on using other assessments, such as attempts or active plans as outcomes?
GS: Many studies use a number of suicide-related events as outcomes, including attempts, ideation, ideation with active plans, ER or inpatient visits due to ideation and/or attempts, and non-suicidal self-injury. Some studies create a suicide composite as their outcome (including some of the events listed above), and might look both at the composite as well as the individual behaviors in their analyses.
PRIM&R thanks Dr. Siegel and Dr. Stanley for sharing their expertise.
The recording of this webinar is available for individuals to purchase in PRIM&R’s online store. If you would like to purchase the webinar for group viewing, please download the order form (PDF) and send it to email@example.com.
*No endorsement of PRIM&R, its programs, or services, by NIH is intended.